metadata

task_categories:
  - summarization

AutoTrain Dataset for project: musicprompt

Dataset Description

This dataset has been automatically processed by AutoTrain for project musicprompt.

Languages

The BCP-47 code for the dataset's language is unk.

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "text": "['instrumental', 'medium tempo', 'electric guitar lead', 'ambient', 'steady drumming', 'groovy bass line', 'trumpets', 'melodic', 'pleasant', 'funky', 'groovy', 'soft rock', 'pop rock', 'funk rock', 'youthful', 'atmospheric', 'brass band', 'soul', 'neo soul', 'soothing', 'rhythmic acoustic guitar']",
    "target": "This music is a melodic instrumental. The tempo is medium  with a captivating electric guitar lead, rhythmic acoustic guitar, funky bass line, keyboard accompaniment, steady drumming and trumpets. The music is soothing, atmospheric, euphonious, youthful, and soulful. This instrumental is a Soft Rock/Funk pop."
  },
  {
    "text": "['pianomusic/meditation', 'water soundsample', 'acoustic piano', 'reverb']",
    "target": "This song contains a piano-composition with a lot of reverb playing a relaxing melody while running a waterdrippling sample. This song may be playing at home for meditation or sleeping."
  }
]

Dataset Fields

The dataset has the following fields (also called "features"):

{
  "text": "Value(dtype='string', id=None)",
  "target": "Value(dtype='string', id=None)"
}

Dataset Splits

This dataset is split into a train and validation split. The split sizes are as follow:

Split name	Num samples
train	2159
valid	540